Linguistic Teaching Decomposition

Poetry Translation with a View of Language Learning
(To Have, or Not to Have?)

Traditionally, translation is the interpretation of the meaning of a text in one language (the „source text“) and the production, in another language (the „target language“), of an equivalent text (the „target text,“ or „translation“) that communicates the same message.

The translation has to take into consideration various constraints including the context of the source text, idioms, grammar rules of both languages, writing conventions of the target text and perhaps many other things that may be taken into consideration.

But sometimes constraints become too difficult and contradictory to be implemented. For example, poetry is very difficultly (if not at all) translatable due to conveying both the form and the matter from the source language to the target one. In 1959 the well-known linguist Roman Jakobson declared the poetry untranslatable on the pages of his paper „On Linguistic Aspects of Translation.“; the same point of view was supported by the American poet James Merrill, who wrote his poem „Lost in Translation“ in 1974. The question was also considered in Douglas Hofstadter’s 1997 book, „Le Ton beau de Marot“.

The sung text translation (or singing translation) is even more restrictive than the „untranslatable“ poetry due to a European custom to set the sung text to a verse. Really, it is very difficult (if not possible at all) to add music to the translated sung text taking into consideration the problem of original syllables fixed to the sounds of a melody. The syllables of the translated text (even if you would be able to receive the translated text) often do not properly correspond to the melody which is far more natural to the source language.

In addition, it has to be said that the poetry and songs have its imageries which are very difficult to preserve in the translation. That is why; very often poetry translations and especially song translations convey distorted imageries or no source imageries at all. This explains a problematical attitude to any translated poetry and songs.

Well, but if you are not able to translate the poetry correctly, whether you can have any use of the poetry with a view of the foreign language learning? My answer is positive: YES, YOU CAN HAVE! People always had been using poetry and songs before modern times, they use them nowadays and people will always be using such or similar things in the future.

But the memorization of lyrics and verses means too little for learning a foreign language.The understanding of a poetry is also insufficient for you (in this case the text translation can help you to see the message of verses), but you have to know language means, with which the poetry translation was carried out. In the last case the translation is of no use for you. This problem can not have simple decisions. Usually in similar cases, you are recommended to consult a competent teacher.

The VStoR program is targeted for self-studying the genuine Russian songs. These songs were not created for teaching a foreign language. The lyrics of the songs convey real Russian, and they have all aspects of a living language.

The special method, termed Linguistic Teaching Decomposition (LD) has been developed to make your self-learning really useful and effective. This method has been created for application in learning lyrics. You can read about LD principles in the following text.

Our Solution

Linguistic Teaching Decomposition (LD) is the method of multilevel analysis of an original poetic text with the help of an intermediate language.

The task of the LD method consists in finding out and in the studying of the linguistic means with which the original has been written.

By employing the LD method you can (the purposes of the method):
  • actively enlarge your vocabulary;
  • acquire and improve your skills in using Russian words and phrases in specific speech situations;
  • really be helped to perceive a song in the Russian language.

Brief Explanations and Definitions

Source language is the language, in which the studied poetry is written, and this is the language, which is under study. In the LD method, the target language coincides with the source language. This fact indicates an obvious distinction from traditional translation, in which the target language is the language of translation. (This definition and its explanation is very important for understanding the LD method!)

Original text is the source text, written in the source language.

Assistant language (AL) is the intermediate language, which is supposed to be clear to the user. The teaching material is written in AL with a view of the LD method. The teaching material structure and content exceeds the limits of the traditional translation. For this reason we do not use the term of the target language, since the purpose of the Linguistic Teaching Decomposition method is not at all the translation, but the studying of the source language.

Direct translation is a literal translation bearing in mind the context of a specific sentence.


There is some process, contained inside the realization of the LDmethod. This process has a certain similarity to the mental  process, which goes in a mind of a translator while producing a text translation.

The reader of a translated text has no interest in this process, but the same process has a great value to the person studying the language of an original.

Since you need to perceive a song in a source language, the LD method does the sequential analysis of a text at several levels, using the assistant language. After this analysis, the user returns to analyzing and understanding the source text.

In the fig.1, you can see the schematic representation of LD levels. In the fig.2, you can establish the correspondence between LD levels and elements of the user interface.

Fig. 1. Levels of Linguistic Teaching Decomposition.


Fig. 2. The LD method realization in the "Via Songs to Russian" software.

LD Levels Description

1. Vocabulary Level

Usually, this is a fragment of lyrics from a teaching sentence. The sentence is split into separate words (lexemes); each word can be identified by the electronic dictionary. As a matter of fact this is a lexicon, which the user can memorize, while working with the song.

The note: There are two panes, which correspond to the vocabulary level. They are the source text pane and the dictionary entry pane.

2. Decomposition Level

At this level the source sentence is split into fragments. Each fragment consists of a single word or several words. This fragment is called as a decomposition item.

The splitting is implemented with a view of a received decomposition item minimization. Each original decomposition item has to have the correspondent decomposition item in the assistant language, which is to be „as close as possible“ to the original decomposition item. Such a decomposition item (a fragment of a sentence) constitutes any concept, type of locution, or figurative turn of speech.

The note: The „as close as possible“ phrase is used because sometimes you can not find direct conformity to some words or concepts in the lexicon of other language. In this case, any adequate direct translation becomes pointless at all.

If some  decomposition item has details about a song subject, which is not to be distorted, or lost, then the content of such a decomposition item is called as a subject element.

The subject element is a key notion, which is to be perceived by the user. Using this notion you can establish a bridge between two languages: the source language and the assistant language.

Sometimes, the decomposition item has no significant information about the song subject. In this case it remains at the binding level (the 3rd level, or a direct translation) of the LD method. This item can be lost at 4th level (the subject level). This fact will be described further during definitions of correspondence types.

In some cases basic syntactic differences of languages does not allow to find out the decomposition items of reasonable size. Often, this situation can be resolved with the adaptation of the source sentence. The adaptation is implemented with the alteration of a sentence word order to the sentence word order of the assistant language syntax. This implementation is to be done without any violations of grammar rules, or a sentence message.

Often, the adaptation of a source sentence is carried out for the visualization and simplification of perception by the user. In such a case the word order of a source sentence is adapted to the word order of the assistant language.

In the fig. 2, you can see the example of such an adaptation. Here „чёрный кот“ words are shifted to the beginning of the sentence. Such a word order is typical for the English language, which is the assistant language here.

In fig. 3, you can see the example the sentence adaptation with a view of decomposition items optimization.

Fig. 3. Here, you can see the adaptation example of the Russian sentence to the syntactical word order of the English sentence.

In the user interface, the rearranged word is pointed at the left-hand with the green arrow directed inside, while the place of its actual position is designated with the red arrow directed outside, and the word is outlined with the grey color, while being crossed out.

The note: In most cases, the Russian sentence syntax allows the adaptation. The sentence adaptation has a practical significance for the user, since most of the adopted sentences can be easily decomposed and understood in the assistant language.

3. Binding Level

Usually, the translator does not aspire to draw a parallel between the source and translated text, since the reader does not have necessity to compare these texts with a view of their structural correspondence. The task of translation is to render into another language the message expressed in an original language, instead of saving the structure of the source text.

If you learn any language, the reading of translations is ineffective, since the translated text can differ greatly from the original text. The translated text can have a weak correlation with the structure of an original. Certainly, we do not consider the case of teaching texts, which are specially written for tasks of foreign language learning.

If you study genuine songs, the matter is quite different. The genuine songs have a lot of various locutions, figurative turns of speech. Even high professionals have great difficulties in song translations, but often the translations are impossible to achieve at all. In some cases it is possible to explain only the message of untranslatable words and expressions in a free-form manner. Quite often similar interpretations can be verbose and rather obscure, and can do little good for teaching purposes.

This level is introduced to have a direct correspondence between the lexicon of studied language and the lexicon of the assistant language. This level carries out the task of placement in the user’s memory of some kind of „anchors“, or of „labels“, with which the new information from the following level can be associated.

At this level, there are translations of decomposition items, for which the following two rules can be applied:

  • the word translation can be selected only from the software dictionary;
  • if the word has several meanings, the most appropriate one is to be selected.

For example, the „коса“ Russian word has several meanings; the two of them are a „plait“ and a „scythe“. If the context is about female hairs, the „plait“ term is to be used. If the context is about haying, the „scythe“ term has to be used.

Important! Though, this translation is the direct one, the context is taken into consideration. Therefore, the Binding Level translation can be called as a direct contextual translation.

4. Subject Level

In most cases the direct translation (the binding level translation) distorts the author’s message in lyrics. Often, the distortions appear because of psychological and cultural distinctions between native speakers of two languages.

The task of this level is to restore the subject of the song which is under study, with the help of the assistant language. The translation is to convey concepts and ideas (subject elements) of the source text. At this level, the decomposition item, which contains one, or several subject elements, becomes a grammatically and stylistically correct fragment.

Actually, for the concept expressed with a turn of speech in the source language, the appropriate turn of speech most closely describing this concept in the assistant language is found out and put into a correspondence.

Important! Even absolutely correct translation of the subject element (from the point of view of grammar and stylistics of the assistant language) can not guarantee the clear and unambiguous disclosure of its message (in this case the next (5th) level is used).

The following example can explain the above note. Imagine yourself among theater audience, watching the scene show. On the scene, you can see the walking man. Suddenly, a black cat has crossed his path. The man stops, crosses himself and then chooses another path.

The translation of this episode can be implemented without problems, but audience, who are not familiar with the Russian bad omen „about a black cat“, would not understand the episode. The similar situation, if a child watches an adult film. The child can see and hear everything perfectly, but it has little understanding of this film.

Types of correspondence

At the subject level (the 4th level), the decomposition item has to be grammatically and stylistically normal (in accordance with the assistant language). Often this fact is the cause of an essential distinction between the direct translation and the subject translation. Sometimes, the distinction is far greater, and the subject translation has almost nothing in common with the direct translation.

To get at the root of this distinction, there is the concept in the LD method, which is called a type of correspondence.

The type of correspondence is established for the decomposition item between its direct and subject translation, and it has 5 variants listed below:

 no correspondence.

  Direct — the type of correspondence when the translation of the binding level completely coincides with the translation of the subject level. Perhaps, this is the most simple and easily comprehensible type of correspondence. (Really, it agrees with the literal translation)

Approximate — the type of correspondence when the translation of the binding level needs grammatical correction to become the translation of the subject level. The distinction between the direct and subject translation is not great.

Below, you can find situations, in which the approximate correspondence is permissible (English is the assistant language, the source language is Russian):

  • adding/deleting of various syntactic words according to the rules of the English language;
  • adding of pronouns, if the transformation of an impersonal phrase to the personal one is needed;
  • applying of morphological modification of words, if the sentence syntax demands to do it at the subject level.

Fig. 4. Here is the multiple example of the approximate type of correspondence.

Equivalent — the type of correspondence, when a paraphrase is used. This type of correspondence is widely used. The result of its application can be the stylistic correction, or the complete replacement of the directly translated phrase.

The equivalent type of correspondence can be applied in two basic situations:

  • The direct translation of a phrase is stylistically incorrect, and, as a rule, it can not be perceived without difficulties. Perhaps, this phrase can be perceived by someone, but the same phrase can be written in recognizable form, having the same meaning as the source phrase.

    You can see the example of a similar situation in fig. 2. The „жил да был“ phrase is the quite ordinary one for any Russian speaker. Literally, this phrase can be translated like „lived and did exist“, but at the subject level it can be translated like „once lived“, or „there was once“, that is more natural to English.
  • Saving the direct translation of a locution onto the subject level is impossible because of the complete or partial loss of its sense. It happens, if the same locutions are expressed in two languages with different means, the source locution can not be translated literally. But it should be pointed out, that in the case of the equivalent correspondence the source locution bears no figurativeness.

    Let’s look at the following example. In the „Katyusha“ song, the subject translation of the „Пусть он вспомнит девушку простую“ clause can be as „Let him remember the open-hearted girl“. The direct translation of the „девушку простую“ phrase as „simple girl“ — obviously distorts the message, which is well clear to Russian speakers. Therefore at subject level this phrase is translated as „open-hearted girl“.

Fig. 5.Here is the distortion of the message because of distinctions in the use of words in different languages.

Sharpening — the type of correspondence, which is "easy" semantic correction. It is applied in cases when the direct translation is admissible, but there is a word or a group of several words in the assistant language expressing more precisely the source concept.

The result of such a correction is either replacement of any word with another one, or adding of a new word (or rarely of several words) specifying the sense of a subject element.

Example 1

Example 2

Fig. 6. Here, you can see the examples of the specification of meaning.

Figurative — the type of correspondence, which is used in the case of figurative expression being in the source text. Proverbs, idioms, popular expressions, sayings, various types of trope — all relate to figurative expressions. In such a situation the direct translation of the decomposition item results in empty words having no semantic connection. In other words, the direct translation leads to the complete loss of message.

At the subject level, the decomposition item is the locution reflecting the precise message of the source phrase. But this locution bears no figurativeness in the assistant language.

The poetry (especially lyrics) very often uses figurativeness to minimize the size of text and to make the text livelier.

Fig. 7 Here, you can see examples of the figurative type of correspondence.

No correspondence — the type of correspondence, when the decomposition item bears no subject element, and it is used as an auxiliary means (for the formation of rhyme, or of verse meter), or it is an untranslatable specific pattern of the original language.

You can find such a decomposition item at the biding level, but it can not be saved to the subject level.

Fig. 8. The "и" conjunction is an auxiliary means for the meter creation of this verse — lost at the subject level.

Important! The point presented to you on the subject level is a possibility to memorize and to know how to use words and phrases in specific speech situations in a studied language.

5. Composition Level

Here, you can observe the assemblage of decomposition items of the subject level into a syntactically correct sentence in the assistant language. Sometimes, authors term such a process as normalization. It is possible to consider the received sentence as „symbolical“ translation in traditional sense of the „translation“ word.

Important! We use the „symbolical“ word, for the reason, that the poetry is considered as mostly being untranslatable. Therefore, you have to learn to perceive the poetry in the source language, while the LD method is simply used as a help for you to study the source text.

6. Semantic Level

As it has been pointed out above, if we see something is happening before our eyes, it does not mean that we understand it. Often, compact poetic patterns bear subtle semantic shades, which can be identified only by speakers of a native language. But frequently even the native language speakers argue on the subject of song understanding, i. e. the same song fragments evoke different imageries in people minds.

The semantic level is not the mandatory one. The level is used by the person, who produces a teaching material when it is required to explain to the user specific semantic nuances, exceeding the limits of the composition level. These may be the features of usage of any words or phrases in informal speech, be the message of a phraseological locution which does not have analogs in the assistant language, any national features linked to some concepts and so on.

Such an explanation can be written in a free form and is tearmed as the Semantic commentary.


By using Linguistic Teaching Decomposition, you can learn consistently all levels of the analysis. No doubt, that your perception of lyrics after the analysis improves considerably. However, the LD method is not a simple procedure for perception and understanding.

Do not fatigue yourself with learning of LD successively many a time. Even if you don't understand everything, choose another software mode or subroutine, which is more simple and interesting from your point of view.

The deep understanding of a song can not come at once, but your understanding is formed step-by-step. Return to the lyrics analysis via certain stages of song mastering: for example, after having worked for some time in the Singing mode, or after having a good time in the Tutor. Do not hurry up to comprehend lyrics at once.

Undoubtedly, you will acquire the perception of any Russian song. But in the process of the comprehension of a song, you receive knowledge unforgettable for the rest of your life, and this is even more important!